Association for Public Policy Analysis & Management

This paper investigates the effect of biased and stereotypical beliefs of teachers on learning outcomes of secondary school students in India. Using the Young Lives school survey data from 2016-17, we measure teacher’s bias through an index derived out of various questions on teacher’s subjective beliefs about the role of gender and other characteristics in academic performance. We tackle the potential endogeneity of teacher’s subjective beliefs by including teacher fixed effects in a value-added model that controls for lagged cognitive ability of students. First, we document that girls perform worse than boys in math and this gap increases over an academic year. Then, our estimate shows that a standard deviation increase in biased attitude of the math teacher widens the gender gap in students’ math performance by 0.07 standard deviation. Furthermore, only male teachers who are biased have a harmful effect on girls’ math performance relative to boys, while female teachers do not have any such negative effect on girls. Moreover, among the medium-performing students, having a female teacher significantly reduces the gender gap in math performance. As plausible mechanisms, we show that biased teachers also negatively affect girls’ attitude towards math as compared to boys. Unlike math outcome, we do not find any significant effect when we analyze the effect of biased English teachers on English scores of the same students.